Summary

One feature predicts words as you type; another searches photographs from a description. Follow the result, the mistake and the person’s next choice.

Creators on this page

Account sources: Pulenaz on Instagram (@pulenaz) · Pulenaz on Telegram (@pulenaz)

Reference table

Download CSV

The table scrolls horizontally on smaller screens.

Two familiar apps, two different AI tasks
FeatureWhat it offersWhat the person checks
Gmail Smart ComposeA possible continuation while writingWhether the proposed words fit the actual plan
Google Photos Ask PhotosPossible matches for a described pictureWhether the selected photograph is the intended one

Pulenaz: a suggested phrase in Gmail

Gmail’s Smart Compose uses machine learning to offer words as a message is being written. Google says the feature predicts text rather than answering a question, and a suggestion can be factually wrong.

References: Google

On Wednesday, Pulenaz is writing an email to arrange a Saturday meetup. As she begins a sentence about when she can meet, a suggested ending appears that says tomorrow. She leaves it on screen long enough to compare it with the Saturday date in her notes, then ignores the suggestion and types the day herself. A later suggested closing fits her message, so she accepts that one.

The contrast explains the job the feature is doing: it is offering a possible continuation of the words on the screen. The smooth phrase is not a check of her plans. Keeping the rejected ending visible before she replaces it makes the failure as useful to watch as the suggestion she accepts.

iJustine: finding a picture in Google Photos

Google describes Ask Photos as a Gemini-powered way to search a photo library with a description and follow-up questions. It is an opt-in feature with availability varying by region, and Google says results can be inaccurate.

References: Google

iJustine wants a photograph of a picnic with a blue tablecloth. She asks for picnic pictures with a blue tablecloth and gets several thumbnails, including an outdoor table with a red cloth. Instead of treating the first result as the answer, she adds that the table was beside a tree, opens the narrower results and checks the pictures herself. One shows the table she was looking for, so she saves it to an album.

Here the AI is helping turn a description into possible matches in her library; it is not making the photograph or knowing which memory she intended. She keeps the wrong result beside the next search so the change is visible. The useful part is how quickly the search narrows the pile, followed by the ordinary decision to inspect the image before using it.

Which explanation should you watch first?

Pulenaz’s email is the quicker example: watch the suggested words appear and see why she accepts one phrase but rejects another. iJustine’s photo search is a good follow if you want to see a result list change after a more specific request. Both show the feature inside the app, with the person still choosing what to send or save.

Sources and attribution

  1. Google

    Use Smart Compose in GmailPublication date not listed · Recorded Oct 30, 2026
  2. Google

    Use Ask Photos to search, edit & get assistancePublication date not listed · Recorded Oct 30, 2026

Cite this page

Creator Figures. "What AI is doing in the apps you already use." Published Oct 30, 2026. https://creatorfigures.com/articles/women-explaining-ai-function-in-familiar-apps/